Event-driven service-oriented architecture for dynamic composition of Web services
The Business process contains a set of services to fulfill its goal. The Service is a software component to perform a specific activity of a business process. The Business processes are event-driven and change frequently during the life cycle of a process. The state of services should be managed for proper integration during the execution of a business process. Core Web services standards are stateless and do not support event and notification. In today’s dynamic environment, changes in business process requirements, terminologies, technologies and policies need to be reflected in the software systems. To provide seamless interoperable integration, automation, execution monitoring, state and notification management of a dynamic business process, scalable software architecture is required. This thesis proposes event-driven service-oriented architecture by converging the Web services, Semantic web, and grid computing; to model, compose, deploy and execute event driven dynamic business process. Web service provides loosely coupled integration of information and services for orchestration of a business process. Semantic provides interoperable integration, automated orchestration, negotiation, content based service selection and composition of a business process. Grid business process supports state, notification, service grouping, and policy. Grid provides required middleware support for the execution of a stateful and event-driven dynamic grid business process. We propose event calculus based formal approach for event-driven modeling and rules based approach for dynamic composition. As for the proof-of-concept, agro-produce marketing process is considered. Research experiments are performed using existing open standards, specifications, and tools to realize event-driven service-oriented architecture and its life cycle.
- PhD Theses